The Tragedy of the Tropics:
A dynamic, cross-scale analysis of
deforestation incentives
by
Bernardo Strassburg
CSERGE Working Paper EDM 07-02
The Tragedy of the Tropics:
A dynamic, cross-scale analysis of deforestation incentives
by
Bernardo Strassburg
Centre for Social and Economic Research on the Global Environment (CSERGE),
School of Environmental Sciences, University of East Anglia,
Norwich, NR4 7TJ, United Kingdom
Author contact details:
Tel: ++44 (0) 1603 591346
Fax: ++44 (0) 1603 593739
Email: [email protected]
Acknowledgements
The support of the Economic and Social Research Council (ESRC) is gratefully
acknowledged. This work was part of the interdisciplinary research programme of the ESRC
Research Centre for Social and Economic Research on the Global Environment (CSERGE).
ISSN 0967-8875
Abstract
Tropical deforestation is the second largest source of anthropogenic GHG emissions and the
major driver of the ongoing wave of mass species extinction. Yet, despite strong
international pressure and diverse initiatives by tropical countries, the conversion rate has
remained unabated for more than 25 years. Here we introduce a dynamic marginal approach
that analyses the distribution of conservation benefits and costs across geopolitical scales
and their behavior as deforestation advances. The case study focused on the Brazilian
Amazon projects that under the current regime of incentives the conversion will continue
until 70% of the original forest is gone. Despite a few quantitative uncertainties, this
approach provides insights for future research and points to a systematic international
cooperation as the key to halt tropical deforestation.
INTRODUCTION
After being historically concentrated in the temperate regions of the North1, the conversion
process is now focused on the most diverse and productive ecosystems of the planet, the
tropical forests. Despite its 1.3 billion hectares4 representing only c. 8.6% of Earth’s
terrestrial surface, tropical forests produce 31,6% of annual land NPP5, and hold between
half and two thirds of all living species of the planet6.
It is estimated that between 8.1[7] and 15.2[4] million hectares of tropical forests are converted
every year. Resulting greenhouse gas emissions amounted to 18-25% of global
anthropogenic emissions8,9, making tropical deforestation the second largest source of
emissions, closely behind Energy (24%) and well ahead of Transport (14%) and Industry
(14%)8,9.
The conversion of natural environments is the main driving force behind the explosive
growth in extinctions rates10, presently between a hundred and a thousand times higher than
natural rates, with a further tenfold increase expected in the near future1. Recent field
observations11 appear to support these catastrophic scenarios.
Not surprisingly, the conversion process is considered1 the largest negative impact factor in
the provision of ecosystem services, the purely anthropocentric concept encompassing the
benefits nature provides to mankind12. Yet, despite strong international pressure and several
and diverse attempts made by their host countries in recent years, the rate of conversion of
tropical forests remains staggeringly high.
The impact of the projected massive loss of Earth’s biodiversity has been more completely
understood in the recent past, when the main significance of biodiversity changed from the
potential value of individual species to their role in the maintenance of ecosystem functioning
and, consequently, in the provision of ecosystem services.
This paper investigates how the distribution of ecosystem benefits across different
stakeholders on different scales has a fundamental role in the persistence of tropical
deforestation and the consequent GHG emissions. An initial systematic analysis of case
studies available in the literature identifies some interesting patterns and suggests a decisive
role for one of the classical failures of economic systems. Following, a novel approach
applied to the conversion process in the Brazilian Amazon aims to test this hypothesis. Even
if some of the quantitative results should be regarded cautiously, this approach offers
important insights and applications for both research and policy.
Its strong qualitative conclusions suggest that if we are interested in safeguarding what
these forests represent for our own wellbeing, it will be necessary to cooperate
internationally. If implemented, such cooperation would be much cheaper, straightforward
and beneficial than recent international initiatives such as the Kyoto protocol. Far from being
competitive, however, both approaches are complementary and strongly synergistic.
THE BIASED TRADE-OFF
As converting an ecosystem has, on the one hand, a significant negative impact on the
services supplied by it to humanity, while on the other hand it increases the production of
alternative valuable goods, there is a clear trade off in the choice of conserving or converting
a given area. The main characteristic of this trade off is that it is strongly biased in favor of
the option for conversion. There is ample literature showing how our growing but still limited
knowledge of the workings of environmental systems, in conjunction with well-known failures
1
in the economic system that are especially evident in this context, create a strong bias
against the conservation option for a given natural environment 1.
An important portion of this bias derives from the fact that while the benefits provided by
converting a given natural environment to alternative uses are well known, observable and
measurable, the benefits provided by conserving the environment in its original form are still
imperfectly understood and their measurement extremely complicated due to the slight and
flawed insertion of these benefits into the economic system. Even when they are adequately
identified, environmental services are usually measured either in physical units (such as
volumes of regulated atmospheric gasses) or in more abstract units (such as cultural
values), which are normally undervalued in comparison with the immediately observable
monetary values of the benefits of conversion.
The valuation process is an important tool in attenuating the bias mentioned above, although
still subject to important and much debated limitations2. Its objective here is to make an
adequate identification of the total economic value of the natural environment in question,
leading to a less flawed comparison between the benefits of a natural environment and
alternative land-uses.
ANALYSIS OF CASE-STUDIES
As converting an ecosystem has, on the one hand, a significant negative impact on the
services supplied by it to humanity, while on the other hand it increases the production of
alternative valuable goods, there is a clear trade off in the choice of conserving or converting
a given area. The academic literature is rich in works that apply valuation methods to
estimate the diverse benefits provided by natural systems and their components. Much less
frequent, however, are works that make comparisons between the benefits of conservation
and those of conversion to alternative uses. After a review of the literature, 14 works were
selected that compare, directly or indirectly, the benefits of these two alternatives (Table 1).
From a systematic analysis of these works, two important observations can be made. The
first is that in most cases – in eleven of the fourteen works – the conservation option
presents superior benefits than the conversion to alternative uses. The three works14,20,21
that do not reach the same conclusion analyze ecosystem benefits only of a local character,
and, as explicitly noted in one of them21, the inclusion of benefits on other scales would
probably reverse the conclusion.
1
Pearce (1991), Barbier et al. (1994), Pearce (1994), Turner et al. (1994), Hempel (1996), Jakobsson and
Dragun (1996), Manson (1996), OECD (1996), Swanson (1996), Swanson (1997) and Fearnside (2000) and.
2
See, for example, Costanza and Folke (1997), Rees (1998), Daily et al. (2000), Turner (2000)and Farber et al.
(2002).
2
Table 1.1:
Country
Region
Analysis of Case Studies
/
Ecosystem
Brazil13
Trop. Forest
Peru14
Trop. Forest
Cameroon15 Trop. Forest
Malaysia16
Trop. Forest
Brazil17
Trop. Forest
Brazil18
Trop. Forest
Madagascar
19
Trop. Forest
Sri Lanka20 Trop. Forest
Kenya21
Savanna
Thailand22
Mangrove
United
Kingdom23
Temp.
Forest
Mixed
Australia24
(Kakadu)
Temp.
Canada25
Wetlands
Scandinavia Boreal
26
Forest
Alternative
Conclusions
Use
Agric./Cattl
Ecos. Glob. Value > $ Conversion
e + Timber
Agriculture
$ Conversion > Ecos.
+ Timber Value
Ecos. Glob. Value > $ Conversion > Ecos.
Agriculture
Value
Ecos. Glob. Value > $ Conversion ≈ Ecos.
Timber
Value
Ecos. Glob. Value > $ Conversion > Ecos.
Cattle
Value
Ecos. Glob. Value > $ Conversion > Ecos.
Average
Value
Timber + Ecos. Glob. Value > $ Conversion > Ecos.
Agriculture Value
Timber +
$ Conversion > Ecos.
Agriculture Value
$ Conversion > Ecos.
Average
Value
Shrimp
$ Conversion < Ecos.
Farm
Value
$ Conversion < Ecos.
Timber
Value
$ Conversion < Ecos.
Mining
Value
$ Conversion < Ecos.
Agriculture
Value
$ Conversion < Ecos.
Agriculture
Value
Nat.
Nat.
Nat.
Nat.
Nat.
Nat.
Nat.
Nat.
Nat.
Nat.
Nat.
Nat.
Nat.
The second conclusion is reached with the introduction of the different spatial dimensions of
the ecosystem services in the analysis. While some services have a local or regional
character, other services have a global dimension. Dissociation and classification of benefits
for each service according to its spatial character demonstrates a highly relevant pattern for
understanding the behaviour of the conversion process. On the one hand, it is observed that
in all four works that analyze natural environments in high-income countries23-26,
conservation shows superior benefits, even when only benefits of services of a local or
national character are considered.
On the other hand, of the nine studies that analyze national or local ecosystem benefits in
medium or low income countries14-22, only the study related to mangrove ecosystems22
reaches a similar conclusion. Of the other eight, one is not conclusive in this respect16 and
the other seven reach exactly the opposite conclusion, that is, that the national benefits of
the natural area are inferior to the benefits of conversion to alternative uses. When the global
benefits are included in the analysis, however, the balance always tilts towards conservation.
This second observation demonstrates the fundamental role played by a well-known failure
of the economic system – that is externality. An externality is the consequence (positive or
negative) that an action performed by a group of stakeholders has on another group that
does not take part in the decision process. Its occurrence in this case is related, on the one
3
hand, to the different spatial dimensions of the services provided by the natural
environments, and on the other, to the organization of the global geopolitical system in
sovereign and independent nation states. The consequence of these two factors together is
that a portion of the benefits provided by the ecosystems is appropriated by individuals in
other countries that do not participate in the internal decision-making process of the
countries that own the resources.
A KEY MARKET-FAILURE
The role of the externality can be better understood in Figure 1. The lower (or leftmost) curve
represents the marginal benefits provided by the natural area to the owner of the land. When
we add the local or regional benefits provided by the natural area to the citizens of the
country (the national externality), the second curve is formed. Finally, the inclusion of the
benefits that affects everyone on Earth (the international externality) leads to the higher (or
rightmost) curve.
Figure 1:
As the remaining natural area diminishes (i.e. when we move to the left along the horizontal
axis), its marginal benefits grows – and grows faster. This expected shape, widely published
in the literature27-31, derives from two fundamental causes. First, basic economic principles,
such as the law of diminishing marginal utility indicate that unit values of the goods or
services rise as their available quantity diminishes28-31. Second, ecological principles, such
as the existence of thresholds and the resulting discontinuities indicate a rise in the
possibility of rupture in the ecological system as larger portions of the area are converted,
putting the stability and resiliency of the system at risk29-34. Thus, it is expected that the curve
representing the marginal unit value of the environmental benefits becomes less elastic
(more inclined) as conversion progresses.
An optimum or equilibrium point occurs when marginal benefits equate marginal costs, the
latter here composed of management and, mainly, opportunity costs (i.e. the forgone
benefits of conversion to alternative activities). The presence of externalities leads to three
4
different equilibrium points. As can be seen, more land should be conserved from a global
standpoint (Qglob) than from a national one (Qnat).
The presence of the international externality is not relevant in this context when the national
benefits of the natural area – the ones that are (or should be) considered by the decision
makers – are by themselves higher than the marginal cost of conserving it. Graphically, it
means that the present “quantity” of the natural area under analysis is to the left of the point
Qnat. In this case, the optimal decision from the national point of view is to conserve the
area. From the analysis of case studies, this is what happens today in higher income
countries.
The opposite occurs, however, in lower-income countries. In these countries the national
benefits of the natural area are lower than the costs of conserving it. It is only with the
introduction of the portion of the benefits appropriated by the inhabitants of other countries –
the international externality – that the balance tilts towards conservation. Graphically, the
present “quantity” of the natural area under analysis is between points Qnat and Qglob.
This analysis goes along very well with the spatiotemporal behaviour of the conversion
process. Developing countries are marching to the left on the horizontal axis, towards Qnat,
at a rate of 8.1-15.2 million hectares per year4,7. This behaviour is coherent with the
conclusion that their present “Q” is to the right of Qnat.
Many developed countries, on the other hand, after moving to the left for centuries, are now
actually marching to the right, expanding their forest cover1,4. This expansion could be
understood as a result of two main forces: First, having reduced their natural areas up to a
point near the private optimum (Qpv) in the past (ignoring national externalities when state
concern with the environment was low or non-existent), they are now being conducted by
state intervention towards Qnat. Second, as their income grew, their demand curves for
environmental benefits moved to the right, leading to equilibrium points with higher portions
of conserved land.
This analysis allowed the detection of the fundamental role that the international externality
has in the persistence of the conversion of natural environments in lower income countries.
As no mechanism exists today to allow these countries to receive part of the external
benefits generated by their natural environments, they follow the “national optimum” path
previously taken by their richer counterparts and head towards Qnat.
THE DYNAMIC MARGINAL ANALYSIS
The studies analyzed above have a static nature. The values they present correspond to a
“picture” of the conversion process at the moment the studies were made. They allowed
establishing the present position of each group of countries in relation to the equilibrium
points. A better comprehension of the essence of the problem, however, can be achieved by
looking at the “full movie” – a dynamic analysis.
The analysis that follows is an attempt to apply the theoretical approach behind Figure 1 to
the conversion process in the Brazilian Amazon. The analysis will focus on the 84% or the
original forest that still existed in 2002, approximately 336 million hectares35. Since the late
1970s these are converted at a historical rate of 2 million ha-1 yr-1 [35], although a recent and
still controversial study36 suggested that using remote sensing to identify selective logging
would double this figure. Cattle ranching is still the dominant alternative use35, although
cropland (particularly soybean) conversion is on the rise37.
Among the many obstacles encountered, one was by far the most relevant: although the
literature suggests that the benefits of the natural environments have the general dynamic
behaviour described in the previous section, the specific (quantitative) behaviour remains
5
unknown. Several economic and ecological concepts and tools were combined to overcome
this and other obstacles. In some cases the results seem fairly robust, while in others the
estimations are more sensitive to some arbitrary assumptions adopted.
Nonetheless, even taking some of the specific quantitative estimations with reserves, the
dynamic approach adopted offers solid qualitative conclusions, identify a few points invisible
to static studies, raises potential research applications, allows a clearer view of the role
played by the externality and offers insights on how to counteract it.
In order to offer useful insights to policymaking, the values used in cost-benefit analyses
should ideally be of a marginal nature26. In the present context, this means estimating the
value of the benefits that are lost with the conversion of the next hectare of forest and the
costs of conserving it. Due to the commonly used valuation methods and to different scales
of analysis, however, the values normally found in the literature are closer to either
maximum or average values. This subtle but important difference will also be addressed in
the analysis.
An attempt was made to analyze the forest’s Total Economic Value as completely as
possible, based both on original calculations and on the treatment of estimates available in
the literature. This resulted in the analysis of 12 different kinds of benefits provided by the
forest, compared with an average of 4 analyzed in the case studies discussed previously. It
is important to note that despite being a common practice in the literature, analyzing the
benefits provided by ecosystems through individual “services” is a simplification that tends to
ignore important feedbacks38 and complex relations39,40 between the different ecosystem
components, structures and services.
Some of these services were directly estimated, while others were obtained from the
conversion of average values found in the literature. Some of the services commonly listed
in ecosystem services frameworks28,51 do not directly affect human well-being, and should be
considered as prerequisites for other services that have a more direct impact. MEA (2005)1
classifies these services as “support services”. Among them are the services of soil
formation and nutrient cycling. Although the decomposition processes that lead to the
formation of soil in a tropical forest are certainly indispensable for the forest to provide other
services, such as climate regulation, to assign those services a separate value would be
double-counting.
All values were indexed to 2002 US$ using the Consumer Price Index (Bureau of Labor
Statistics: http://stats.bls.gov/cpihome.htm). Values estimated in other countries were
converted to Brazilian values (although still measured in 2002 US$) using the relation
between PPP-adjusted per capita income of both countries.
The services were grouped in four categories corresponding to each of the TEV categories.
The adoption of different approaches to analyze each category has proven to be a key
measure to overcome some of the pitfalls encountered.
DIRECT USE VALUES
These are benefits generated at present by natural systems through mankind’s production
and consumption activities. They are, therefore, the benefits most subject to market
characteristics. The three services included in this group have relatively well-developed
markets. These are services for which there is a limited demand, which is disputed both by
other suppliers of the same goods and services and by other substitute goods and services.
Consequently, the demand curve for them is relatively elastic.
6
In general, studies that estimate direct use values follow a similar process: First an attempt
is made to identify the quantity of each good or service that can be produced in a
sustainable manner within a generic unit (usually a hectare) of the ecosystem under
analysis. Then, the quantity found is multiplied by the market profit of each one of these
products. The values found from this method, however, are in essence maximum unit
values, normally considerably distinct from marginal unit values.
Take the case of forest timber products, for example. The most recent valuation studies of
the Brazilian Amazon18,19 use a value of US$ 33 ha-1 yr-1 as a reference of the benefits
obtained from sustainable timber exploitation of one hectare of Amazon forest. Although it is
perfectly acceptable to affirm that in one hectare of generic forest the quantity of wood to be
produced sustainably can yield the annual return referred to, the use of this value as the
marginal benefit of this activity in the Amazon forest is not correct.
This is due to the fact that all sustainable timber for which there is a potential market can be
produced in an area approximately equal to 10% of the present area of the forest. Therefore,
at the present stage, the next deforested hectare does not diminish the total benefit that
could be obtained from commercial exploitation of sustainable timber. Consequently, at the
present stage, the marginal value of conservation related to timber exploitation is zero. Only
when the remaining area is equal to the area necessary to meet the potential demand does
deforestation of one hectare result in losses from this activity. The same rationale can be
applied to the other benefits corresponding to Direct Use Value.
1.1 Timber Products
Two of the most recent studies reviewed that analyze the benefits of the Brazilian Amazon,
Andersen et al. (2002)18 and Seroa da Motta (2005)17 use the value of US$ 33 per year per
hectare obtained from Almeida and Uhl (1995)52.
According to Almeida and Uhl (1995)52, the volume of timber removed from each hectare in
the 30-year sustainable cycle is equal to 38 m3, that is, approximately 1.25 m3 per year.
Nepstad et al. (1999)53, on the basis of interviews with loggers in the area, present three
different estimates for the volume of timber removed from one hectare of Amazonia. The
lower estimate is 19 m3 for the entire cycle (or 0.63 m3 per year), the medium estimate is 28
m3 for the cycle (or 0.93 m3 per year) and the highest estimate is 40m3 for the cycle (or 1.33
m3 per year).
Asner et al. (2005)36 presented estimates for the years 2000, 2001 and 2002 of respectively
26.6 m3, 21.7m3 and 21.4 m3 per hectare. According to data from IBAMA cited by Lentini et
al. (2005)54, the average volume authorized by IBAMA in 2004 in the Forest Management
Plans was 27.3 m3 per hectare for the 30-year cycle. This volume is also similar to the
average intensity of exploitation on the Small-scale and Community Forest Management
Plans authorized by IBAMA, equal to 0.9m3 per hectare per year (Lentini et al., 2005)54.
The average between Nepstad et al. (199953) medium estimate (28m3), Asner et al. (2005)36
mean (23,2m3) and IBAMA’s FMP authorized volume (27,3m3) is equal to 26m3 for the 30year cycle, or 0,87m3 per year, and will be used in this estimate. Extrapolating for the entire
area under analysis, this estimate corresponds to sustainable production of 293 million cubic
meters of timber per year.
According to IBGE (2006)55, in 2005 the total amount of timber extracted from planted
forests in the Northern Region of Brazil was 4,2 million cubic meters and the amount of
timber in logs “collected” from the forest in the Northern Region was 12,7 million cubic
meters. The total production was thus approximately equal to 17 million cubic meters.
According to Lentini et al. (2005)54, timber production in logs in the Amazon region in 2004
7
was approximately 25 million cubic meters. Once again, the conservative character of the
analysis leads us to use this later figure in this estimate.
Therefore, the annual sustainable yield that would be possible in the region is more than
eleven times greater than the current annual production of timber from Amazonia. Thus, an
extrapolation for the entire area under analysis of the monetary values found in the
examination of one hectare is shown to be a strong misrepresentation of reality.
The area necessary for production of all the timber for which there is a market is much
smaller than the area of the present forest. Therefore, at the present stage, conversion of the
forest does not diminish the total benefit obtained from commercial exploitation of timber.
Consequently, at the present stage, the marginal value per hectare is equal to zero.
When conversion progresses up to the point where the remaining area is equal to the area
necessary to produce the timber for which there is a potential market, each hectare removed
would represent a loss equal to the profit that could be realized from exploitation of the
timber produced on that hectare. The marginal value will then be constant and equal to this
profit3.
Although IBGE (2006)55 foresees that timber production from Amazonia will fall in the next
few years, it will be assumed here that the conservative volume of 25 million cubic meters
chosen will be maintained. A further conservative assumption – in the sense that it increases
the area that would not be converted from the national point of view – will be adopted. The
performance of Brazil in the international tropical timber market is very weak. Brazil is not
among the five main exporters. It is foreseen, however, that the main exporter, Malaysia, will
lose market share with the exhaustion of its extraction areas. It is here assumed that Brazil
will occupy its position, exporting around 5.5 million cubic meters of timber per year (World
Bank, 2000)56. Thus adding the 5.5 million de cubic meters from this new market to the value
found for 2004, the total size of a potential annual market for Amazonian timber would be
equal to 30 million cubic meters.
Applying to this potential market an average intensity of 26m3 for the 30-year cycle, or 0.87
m3 per year, we have the area necessary for sustainable planting to meet the production
desired, that is 34.5 million hectares.
The annual value of US$ 33 ha-1 yr-1, found by Almeida and Uhl (1995)52 – chosen for the
strict methodology and for having been carried out in Brazilian Amazonia – will represent the
marginal benefit of conservation of each hectare below the area estimated above. Thus the
equation that represents the behaviour of the marginal unitary benefits from this service is:
 y = 33 | x ∈ (1;34.000.000)
 y = 0 | x ∈ (34.000.001;336.000.000)
Eq. 1.1 
It is assumed that the profit from exploitation of the timber will be appropriated by national
producers and therefore the values estimated will be added to the national portion of the
benefits of conservation.
1.2 Non-timber Forest products
This refers to non-timber products that can be cultivated sustainably, such as latex, resins
and oil seeds, as well as several types of food. In opposition to the case of timber, there is
3
The legal international market for tropical timber is highly competitive, with suppliers subject to
international equilibrium prices.
8
not enough data available to compare the potential and actual market for NTFPs in physical
units. Therefore monetary values will be used for this comparison, what can lead to distorted
results. As expressed in the limitations sections, however, this distortion would only begin to
have an impact on the analysis carried in this article if the correct minimum area were ten
times greater then the one estimated below. Given the conservative assumptions adopted
here, this is very unlikely to be the case.
According to IBGE (2006)55, the NTFPs production in the nine states that compose the
“Legal Amazon” amounted to US$ 149 million in 2005. It should be noted that this area is
larger than the area under analysis and therefore the potential market will be a conservative
super estimation. For instance, 99% of babacu production (or 34% of total NTFPs value)
comes from the state of Maranhao and from transitional ecosystems different from the
evergreen broadleaf forest that is the focus of this study.
Table 1.2 summarizes the estimates of the potential sustainable production of NTFPs in one
hectare of a tropical forest. As can be seen, the values vary widely, ranging from US$ 11,51
to US$ 1032. As Peters et al. (1989)57 estimate is more than two times greater than the
second largest estimate found and have been considered an outlier before18, it will not be
included in the analysis. The average of the other 9 estimates is US$ 135.91. To be on the
conservative side and to avoid temporal and spatial benefit transfer problems (link ecol
econ), the value presented by Shone and Cavaglia-Harris (2005)62 will be chosen to reflect
the potential economic returns of sustainable NTFPs produced in one hectare. As a further
conservative assumption, instead of using their estimate for potential production (that
included both goods consumed by the families and sold to the market), it will be used the
value referring to the marketed fraction only, or US$ 16.94 ha-1 y-1. Note that this value is
one-eighth the average of the studies found.
Table 1.2:
NTFPs
Study
Peters et al. (1989)57
Anderson et al. (1991)58
Pinedo-Vasquez et al. (1992)14
Chopra (1993)59
Godoy et al. (1993)60
Godoy et al. (1993)60
Grimes et al. (1994)61
Kumari (1994)16
Batogoda et al (2000)20
Shone and cavaglia-harris (2005)62
Mean (without Peters et al., 1989)57
Value Used
Country
Peru
Brazil
Peru
India
Mexico
Peru
Ecuador
Malaysia
Sri Lanka
Brazil
Original Value
422
59
23
98
116
20
105
11
186
34.43
GNP per capita
adjusted
(2002 US$)
636.75
924.16
59.00
77.94
34.70
44.49
285.95
356.10
97.50
121.42
30.18
37.58
215.69
261.76
9.07
10.71
396.81
414.64
34.43
34.43
135.91
16.94
Multiplying this conservative potential market production by the area under analysis, yields a
conservative potential production of US$ 5.7 billions per year. Even this conservative
underestimation of the potential production, however, is approximately 38 times the
conservative super estimation of the market currently existent for NTFPs from the Amazon
region. Therefore, similarly with the timber case, it can be inferred that the next deforested
hectare will not reduce the total benefits from the NTFPs and so the current marginal value
of this activity is still zero.
9
The marginal value will become positive when the conversion process reaches the minimum
area necessary for the production of all NTFPs for which there is a potential market. Dividing
the existent market (US$ 149 million y-1) by the potential monetary returns of one hectare
(US$ 16.94 ha-1 y-1), this minimum area is conservatively estimated at 8.8 million hectares.
Formally:
 y = 16.94 | x ∈ (1;8800000)
 y = 0 | x ∈ (8800001;336000000)
Eq. 1.2 
As in the case of timber products, these values will be added to the national portion of the
benefits of conservation.
1.3 Recreation
The benefits generated by recreational activities in the ecosystem under analysis are related
to the tourism practiced in the region. The values found in the literature for recreational
activities in tropical environments are listed in Table 1.3.
Table 1.3:
Recreation
Study
Tobias and Mendelsohn 199163
Edwards (1991)64
Ruitenbeek (1992)65
Andersen et al (2002)18
Seroa da Motta (2005)17
Country
Costa Rica
Ecuador
Cameroon
Brazil
Brazil
Original Value
52
55
5
1.6
9
GNP per capita
adjusted
(2002 US$)
44.27
58.49
112.98
149.25
17.28
22.16
1.60
1.60
9.00
9.00
The values listed above were obtained through very similar methods, which consisted
basically in measuring or estimating a total value for the market in the tourism sector in a
given region and dividing it by the area of that region.
As was observed by Seroa da Motta (2005)17, “due to the dimensions of the still preserved
area of Amazonia, it is plausible to suppose that the potential for [recreational] activities can
be assured, even with a quantity of forest lower than that currently existing”. Thus, it can be
assumed that the dynamic behaviour of the marginal benefits of this service would be similar
to the services previously analyzed. It would thus be equal to zero in the present stage and
become positive when the minimum area in which the maximum potential per hectare can
still be exploited is reached.
The first three studies listed analyse smaller national reserves or parks in other countries in
which there is a relatively well-developed ecotourism sector, and which could be considered
an estimate of the maximum potential per hectare4 . The average among them is US$ 76.63
and will be used to represent this maximum potential.
The last two studies seek to estimate the maximum potential for tourism in Amazonia. While
Andersen et al. (2002)18 stipulate a value of US$ 800 million per year, Seroa da Motta
(2005)17 assigns to Amazonia a potential per hectare equal to that found today in the region
of the Pantanal, where the tourism industry is much more developed. Multiplying this value
4
The first three studies focus, respectively, on the Monteverde Cloud Forest reserve in Costa Rica, on
the Galapagos National Park in Ecuador and on the Karup National Park in Cameroon.
10
by the area of Amazonia, we find a potential market equal to US$ 3.25 billion. The average
of the two markets is therefore approximately equal to US$ 2 billion a year.
Dividing this market potential by the maximum potential benefit estimated above, we have a
minimum area of 26 million hectares – or approximately 7.7% of the present area of
Amazonia. Considering that below this area greater concentrations of tourist activity would
cause “crowding” and a consequent loss of “the wild experience”, it will be considered that
the marginal benefit from this point on would be equal to the US$ 76.63 ha-1 yr-1 estimated
previously. Its formal representation is therefore:
 y = 76.63 | x ∈ (1;26000000)
 y = 0 | x ∈ (26000001;336000000)
Eq. 1.3 
Even considering that the recreation values from tourism are in part appropriated by foreign
companies and in part appropriated by national companies, in the dynamic costs-benefit
analysis of the conversion process, the value above will be conservatively related exclusively
to the national component.
INDIRECT USE VALUE
These are the benefits provided by the ecosystems today, and which do not depend on
human activity. The main difference from direct use services is that indirect use services do
not depend on markets and are always fully absorbed by humanity. There is no such thing
as excessive water regulation or too much erosion control. This does not mean, however,
that there are no substitutes for these services or ways to compensate a fall in their
provision. The difference is that generally these substitutes are imperfect, expensive and
limited, normally allowing compensation only for relatively small variations on the provision of
the original service. That is another reason why the marginal demand curve of this services
is initially less inclined, becoming steeper as the conversion progress (as shown in Figure 1).
As will become clear, the studies found in the literature concerning 5 of the 6 indirect use
benefits (the exception being climate regulation) usually estimate average values. To carry
out the analysis proposed for this paper, however, it is necessary to use marginal values. It
was thus necessary to convert the average values into marginal ones.
Of all the works reviewed that show similar curves, only one30 proposes a specific function to
describe the behaviour of ecosystem benefits over the progress of conversion. The
proposed function, the rectangular hyperbolic, is used here to represent the dynamic
behaviour of IUVs. As well as having a behaviour similar to that shown in Figure 1, this
function has an interesting application here. As detailed in Appendix A, it allows the average
and total values found in the literature to be converted into marginal values through the
formula:
MgV = AvV / ln(x)
where x is the area of the ecosystem under analysis.
It is important to point out that the choice of the rectangular hyperbolic function is arbitrary
and therefore the marginal values obtained through it should be approached with caution.
However, in addition to respecting the ecological and economic principles cited, it will be
shown that this function generates results that are compatible with the studies of
comparative valuation previously analyzed and with other related studies.
11
2.1 Climate Regulation
The measurement of the global climate regulation service provided by tropical forests is
much facilitated, if compared with the other indirect benefits, by the widespread attention
given to the climatic question that resulted in a large volume of research associated to the
theme and in the creation of carbon markets.
The estimation of the benefits from this service is composed of five stages. In the first, it is
necessary to estimate the quantity of carbon present in one hectare of a tropical forest such
as Amazonia. In the second, to estimate the fraction of carbon that is emitted to the
atmosphere when the forest in converted to other uses. The third step consists of including
the effect of other greenhouse-gas emitted in the process. Then the estimated emissions per
hectare are multiplied by the monetary value corresponding to non-emission of one tonne of
carbon. Finally, as the product obtained is a present value, a discount rate is applied to
obtain annual values.
Estimates for the carbon contained in one hectare of a tropical forest vary widely (Houghton
et al, 2001)66. The value used here is the average of the carbon density used in six studies
focused on the Brazilian Amazon, listed in the first column of Table 2.1, equal to 190 MgC
ha-1. In order to keep the estimate on the conservative side, the release of carbon from soil
during the conversion (varying between 3,92 MgC ha-1[67], 16Mg ha-1[68] and 24 MgC ha-1[69]
will not be added to this estimate.
The next step is to consider the fraction of carbon that is in fact released to the atmosphere
when the forest is converted to alternative uses. Two different approaches are used in the
literature. Fearnside (1997)67 makes a very detailed account of the carbon released in the 10
years following the conversion of a hectare of forest to an “equilibrium landscape”,
containing 4% of farmland, 43,8% of productive pasture, 5,2% of degraded pasture, 2,0% of
secondary forest derived from farmland and 44,9% of secondary forest derived from pasture.
He concludes that this equilibrium landscape would hold 6% of the original carbon biomass.
From the remaining 94%, approximately 42% would be released through three sequential
burnings, 55,6% would be released through decay and 2,4% would become trapped as
charcoal. Therefore, from the initial carbon content in the forest, approximately 91,7% would
be released to the atmosphere.
Achard et al. (2004)70 do not focus on the conversion to alternative uses, but following the
bookkeeping model by Houghton et al. (2000)71, present two estimates for the fractions of
carbon emitted over time in the process of the deforestation: a “best estimate” of 28% (1st
year), 69% (10 years) and 92% (25 years) and a high estimate of 44% (first year), 94% (10
years) and 97% (25 years).
The present estimate will consider that the alternative use will hold Fearnside’s estimate of
6% of the original carbon, or 11,4 MgC. To estimate how much of the remaining 178,6 MgC
will be released, it will be applied to it the average between Achard “best estimate” for the 25
years time horizon (still conservative if compared to IPCC’s 100 years time horizon), or 92%,
and Fearnside’s 97,6%. The resulting estimate is that 94,8% of the 178,6 MgC, or 169.3MgC
(equivalent to 89.1% of the original carbon content)will be released to the atmosphere when
one hectare of the forest is converted. These 169.3MgC correspond to 620.8 MgCO2.
Carbon Dioxide, however, is not the only greenhouse-gas emitted in the conversion process.
Smaller fractions of Methane (CH4) and Nitrous Oxide (N2O) are also emitted. As their
global warming potential is much higher then that of CO2 (21 for NH4 and 310 for N2O),
however, they have a significant impact in the total emissions. Fearnside and Laurance
(2004)72, based on Fearnside (2000)73, suggest that the estimate by Achard (2002)7 should
be multiplied by 1,158 to account for these two greenhouse gas effects. The same will be
done here, resulting in a corrected per hectare emission of 718.9 MgCO2 equivalent.
12
Due to the different methodologies involved, estimates in the literature of the monetary value
of an avoided tonne of CO2 vary widely. Usually, highest estimates look at the potential
damage one extra tonne of CO2 would cause, and lowest estimates look at the mitigation
costs. The mean between the 12 values presented in the third column of table 4 is US$ 7.7
per CO2 tonne. According to Stern (2006)42, the price of allowances in the European Union
Emissions Trading Scheme, the world’s largest greenhouse gas emissions market, has been
in the range of €10 to €25 (US$ 13.3 to US$ 33.3) per tonne of CO2 for most of the period
since it began operating in January 2005. By the end of 2006, however, prices for emission
allowances in the first phase (2006-07) declined sharply, reaching 8.50 (US$ 11.11) in the
end of November74. Prices for the second phase Emissions allowances for the second
trading period (2008-12), remained higher, traded at around 17 (US$ 22.22).
As the values above do not foresee the use of deforestation avoided as an activity that
generates carbon credits, and as this activity would have a considerable weight in this
market and could, therefore, negatively impact the market price, the option is again for a
conservative value. The 718.6 MgCO2 estimated above will be valued at US$1 per tonne of
CO2.
As the consequent US$ 718.6 per hectare is a payment made for the perpetual
immobilization of that carbon, it is a present value. To be compatible with the other values
used in this study, it has to be annualized. Applying an annual rate of 6%, slight below the
average found in the literature, yields the final value of US$ 43.13 ha-1 y-1.
Differently from the other indirect use services, in this case the value has already a marginal
nature, as it corresponds to the benefits of the non-emission of (or to the damage of
emitting) the greenhouse gas from the next deforested hectare.
To simulate the behaviour of this value as the remaining area of the forest diminishes, it will
be used the rectangular hyperbolic function described earlier. Although for the last few
hectares this function projects a very sharp increase in the values, for most of the area (and
for all the area relevant to this study) this choice is also a conservative one. Clarkson and
Deyes (2002)75, for instance, project that the value would rise US$ 1,5 per tonne (c. 5% of
their estimated value) every year, as the GHG concentration in the atmosphere rises. With
the rectangular hyperbolic function adopted here, a similar 5% rise would only occur when
the current area is reduced by 16 million hectares, which would take 8 years according to the
historical rate of conversion. Assuming this rate to remain constant, the value projected by
the rectangular hyperbolic function would be smaller then the value projected using Clarkson
and Deyes (2002)75 5% a year increase for more than 99.9% of the area.
The benefit related to this service is also likely to rise due to an expected increase of forest
biomass in tropical forests in a warmer climate76.
The formal expression of the behaviour of the marginal value of this service is:
Eq. 2.1: y = 14,491,680,000/x | x E (1:336,000,000)
As they are related to a global service, the estimated values will be included in the global
portion of the benefits of conservation.
13
Table 2.1:
Climate Regulation
Carbon Density Carbon
(MgC ha-1)
emitted
192[66]
91.7% [67]
156
[77]
92%
Trace
gas US$
MgC Discount
correction
(2002 US$)
Rate
+15,8% [73]
US$ 4.69 [79] 8 % [84]
[71]
US$ 3.32
[80]
5%
[82]
196 [78]
US$ 24 [81]
2 % [18]
195 [71]
US$ 0.55 [82]
6 % [18]
186 [70]
US$ 2.24 [82]
12 % [18]
217
[67]
US$
20.22 6 %
US$ 29
[75]
[82]
[17]
190
190
89.1%
15,8%
15,8%
US$ 7.7
US$ 1
[79]
US$
443.86 [85]
US$
121.36 [80]
US$
242.72 [84]
US$ 20.18
[82]
US$ 78.48
[82]
US$
706.28 [82]
US$ 20 [18]
US$ 30 [18]
US$ 12 [18]
US$ 18 [17]
US$ 2.73 [18]
US$ 1.37 [18]
US$ 0.27 [18]
US$ 3,6 [83]
US$ 0.82 [17]
Mean
Value
Chosen
Final
Value
US$ 77.94
6,5%
6%
US$ 161
US$ 43.13
2.2 Waste Treatment
Vegetable cover has the properties of storing and recycling certain quantities of organic and
inorganic wastes generated by humanity. Conversion of the forest implies an increase in
damage caused by pollution. Measuring this service, both in physical as in monetary units,
however, is very complex.
The work by Pimentel (1997)86 is based on a physical estimate of the quantity of wastes
produced by humanity and costs of treatment, and estimates a total global value for this
service in the order of US$ 760 billion dollars per year. Costanza et al (1997)28 propose that
this value could be divided by the area of all forests (not just tropical forests) and grasslands
on the planet, resulting in an average value per hectare of US$ 87. Table 2.2 summarizes
the conversion to marginal values. ha-1 yr-1.
Table 2.2:
Waste Treatment
Study
Pimentel et al. 1997 [86]
Original
Value
87
Country
Global
14
GNP
per
capita
(2002
adjusted
US$)
87
99.77
Current
Marginal
Value
(US$ ha-1
yr-1)
5.08
The equation that expresses its marginal behaviour during conversion is:
Eq. 2.2: y = 1,706,880,000/x | x E (1;336,000,000)
Given the local and regional character of this service, this value will be added to the national
portion of benefits of the forest.
Hydrological-related services
The next four services are directly or indirectly related to the hydrological cycle of the forest
and therefore complement each other.
2.3 Disturbance regulation
This service is related to the capacity of the forest to absorb environmental disturbances,
such as storms, floods and droughts. This capacity is fundamentally related to the vegetable
structure of the forest. Conversion of a complex structure such as the tropical forest into a
simple structure such as a plantation or pasture reduces this capacity.
No study was found that estimates the value of this service in the Brazilian Amazon.
Costanza et al. (1997)28 and Torras (2000)29 use the work by Ruitenbeek (1992)65, carried
out in Korup National Park, in Cameroon. The author used the methods of avoided costs to
value this service provided by the local ecosystem, which is also of tropical forest type. The
value found is an average value. So the method described in the beginning of this section
will be used to convert it to marginal value. Table 2.3 summarizes the operations.
Table 2.3:
Disturbance Regulation
Study
Ruitenbeek (1992)65
Country
Cameroon
Original
Value
2
GNP
per
capita
(2002
adjusted
US$)
6.91
10.03
Current
Marginal
Value
0.51
The function that describes its dynamic behaviour is represented below:
Eq. 2.3: y= 171,360,000/x | x E (1;336.000.000)
As it is a service of a local character, this value was added to the national portion of the
forest’s benefits.
2.4 Erosion Control
Vegetable cover impedes soil sediments from being swept away by natural erosion agents,
such as wind and rain. This service has an on-site and an off-site component.
The on-site component keeps the soil nutrients available for plants to use them for their
growth. For the reasons described in the introduction of this section, however, this on-site
component can be considered a prerequisite for providing other services. Adding a value to
this component would be double counting.
The off-site component is related to prevention of silting of rivers as a result of converting the
forest. Maintaining the original cover avoids damage from silting, especially for fishing, river
transport and hydroelectric stations.
15
Seven studies listed in the first column in Table 2.4 estimate the benefits of this service in
tropical forests in different regions of the planet. The fifth column lists the values adjusted to
Brazil and updated to 2002 dollars. According to Torras (2000)29 5, the off-site portion is
equivalent to 5/7 of the total of this service, (sixth column). Finally, these values were
converted to marginal values, listed in the final column.
Table 2.4:
Study
Cruz et
(1988)87
Magrath
Arans
(1989)88
Dixon
Hodgson
(1988)89
Chopra
(1993)59
Chomitz
Kumari
(1996)90
Chomitz
Kumari
(1996)90
Torras
(2000)29
Erosion Control
Country
Original
Value
GNP
per Average
capita
Benefits
adjusted
(2002 US$)
Off-site
Average
Benefits
Off-site
Marginal
Benefits
Phillipines
23
41.05
62.38
44.56
2.27
Indonesia
5
9.29
13.48
9.63
0.49
Philippines
321
585.62
890.01
635.72
32.38
India
145
423.09
526.89
376.35
19.17
Ecuador
4
8.22
9.70
6.93
0.35
Philippines
45
82.10
96.93
69.23
3.53
Brazil
238
238.00
248.69
177.64
9.05
al.
and
and
and
and
The average of these values, equal to US$ 9.60 ha-1 yr-1, will be used to represent the
current marginal value of this benefit. Note that this value is very close to the value adjusted
from Torras (2000)29, which is both the most recent and the only one focused on the
Brazilian Amazon. This indicates that the distortion from temporal and spatial benefit transfer
is not relevant in this case. The equation that regulates the marginal dynamic behaviour for
the value of this service is:
Eq. 2.4: y = 3.225.600.000/x | x E (1; 336.000.000)
As these benefits are of a regional nature, they will be included in the national portion of
benefits of the forest.
2.5 Water regulation
Forest cover exerts a fundamental role in maintaining the hydrological cycle. Differently from
the disturbance regulation service, this service refers to the everyday regulation of the
hydrological cycle and its positive effects on natural irrigation, drainage, regulation of water
discharge into rivers and regulation of river flow.
The studies by Kumari (1994)16 in Malaysia and Fearnside (1997)82 in the Brazilian
Amazonia were updated and converted to marginal value, as shown in Table 2.5.
5
Calculations based on Cruz et al. (1988)87, Dixon and Hodgson (1988)89 and Chopra (1993)59.
16
Table 2.5:
Water Regulation
Study
Kumari (1994)16
Fearnside (1997)82
Country
Malaysia
Brazil
Original
Value
25
19
GNP
per
capita
(2002
adjusted
US$)
20.68
24.42
19.00
21.30
Current
Marginal
Value
1.24
1.08
The average between then is US$ 1.16 ha-1 yr-1 and represents the marginal value of this
service at the current stage of the conversion process. The equation that shows its
behaviour over conversion is:
Eq. 2.5: y = 389.760.000/x | x E (1;336.000.000)
As the service is of a regional character, this value is added to the national portion of
benefits of the forest.
2.6 Water supply
This service is the role of the forest in filtering, retaining and storing water for consumption.
According to FAO (2006)91, Brazil uses only 0.4% of its renewable water resources.
Furthermore, evidence from the literature92 suggests that the role of forests on the provision
of water is ambiguous. There is evidence that a decrease in forest cover in fact increases
the water available for consumption93. For these reasons recent studies have decided to
attribute a zero value for this service18. The positive value chosen here is, therefore, a
conservative assumption.
According to the study by Kumari (1994)16, the market value of water supplied by one
average hectare of tropical forest is equal to US$ 11 . Converting this value to marginal
value yields US$ 0.56 per year for the next deforested hectare.
Table 2.6:
Water supply
Study
Kumari (1994)16
Country
Malaysia
Original
Value
11
GNP
per
capita
(2002
adjusted
US$)
9.33
11.02
Current
Marginal
Value
0.56
The equation that shows the behaviour of the marginal value of this service over conversion
is:
Eq. 2.6: y = 188.160.000/x | x E (1;336.000.000)
As the service is of a regional and local character, this value is added to the national portion
of benefits of the forest.
OPTION VALUE
This service is related to the possible new benefits that can be generated by the forest in the
future, perhaps arising from new environmental, social, economic or technological
conditions. It can be understood as a risk premium that individuals are ready to pay to keep
17
future options open. In the literature, the option value is almost exclusively related to the still
unexploited potential of biodiversity to produce new pharmacological products.
In this case, again, almost all the works available use methods that result in average unit
values. This fact is noted by the only studies found94,95 that actually seek to estimate
marginal values. However, as well as starting from very arbitrary premises, and thus
reaching admittedly debatable results, their methodology is directed to small hotspots of
biodiversity, and cannot be applied to this study. The nine studies found that estimate
average option values for tropical forests are listed in the first column of Table 3. The fifth
column list these average values adjusted for the Brazilian Amazon in the year 2002.
In the same way as with the IUVs, these average values need to be converted into marginal
ones. Differently from the previous section, however, it is possible to reach this objective
without resorting to an arbitrary function. As the benefits in question are fundamentally
related to biological diversity, a function can be reached that relates the marginal option
value of all species in an ecosystem with its area.
A species that has not yet been tested has the potential for development of new products.
The value of this untested species depends on the percentage of successes in previous
tests, on the cost of tests and the profit in case of success. Although some have sought to
estimate this value, in this case this complex operation is not necessary.
Even though every kind of ecosystem has its own values for the constants, the relationship
between an ecosystem area and the total number of species contained in it adapts very well
to a power function, with the general form of:
S = kAz
where S is the number of species, A is the area and k and z are constants. Now, if the total
number of species S is multiplied by the potential monetary value of a species, say w, a
function is reached that relates the total value (TV) of all species as a function of the area:
TV=w.k.Az
or (where c=wk)
TV=c.Az (1)
As :
AV=TV/A (where AV stands for average value):
We have:
AV . A = c . Az
or
c =AV . A/Az (2)
Differentiating (1), we have:
MgV= δtV/δA = z . c . Az-1 (3) (where MgV stands for marginal value)
Substituting (2) on (3):
MgV = z . (AV . A/Az). Az-1 = z .AV . A/Az . Az-1 = z .AV . A1-z . Az-1 =
MgV = z . AV (4)
As z is a known parameter, conversion of the average values available in the literature to
marginal values becomes trivial. The value of the constant z, which defines the inclination of
the species-area curve, is generally between 0.15 and 0.35 [96]. For a large extension of
tropical forest, the average value of 0.25 is a conservative option 97. This value was used in
Eq. 4 to find the marginal values listed in the sixth row.
The average of the values found, equal to US$ 10.26 ha-1 yr-1, reflects the marginal
option value for the current area of the forest. Substituting in equation (3) this value, the
value of z used (z=0.25) and the current area of forest (A=336.000.000), we can identify c,
18
and substituting it in (3), determine the equation that gives the dynamic behaviour of the
option value:
Eq. 2.14 y = 25.462.546 x -0,75 | x E (1: 336.000.000)
As on the one hand the products developed from genetic resources from tropical forests
have a global potential for application, and on the other, the companies that exploit these
products commercially are based in rich countries, the values for this service are normally
considered to be of global character. The same criterion will be adopted in this work.
NON-USE VALUES: BEQUEST VALUE AND EXISTENCE VALUE
Non-use values can be divided into existence value and bequest value. The first represents
the benefit individuals receive from the simple awareness that a natural system (or some of
its components) exists, when no direct or indirect benefits are expected from it in the present
or in the future. The second represents the desire of individuals to maintain the
environmental resource for future generations.
The studies that seek to estimate the non-use value of natural environments usually
estimate the total existence value of the entire system and then divide this value by the area
of the ecosystem, thereby obtaining an average unit value. In this case, again, it is
necessary to convert these values into marginal values through a specific function.
The choice of the species-area curve also seems adequate for this role. A large part of the
studies of contingent valuation seeks to estimate the willingness to pay (WTP) for
preservation of specific species, and the values found are often relatively high. Thus,
considering that species have their own existence value and that the “Amazon Rainforest” is
a system formed by the interaction of its species with the physical environment, it seems
reasonable to suppose that the existence value of the forest is somehow related to the
existence of the species that compose it.
In the specific case of non-use values, this hypothesis can be reinforced in view of the
results from a study of contingent valuation research with citizens of the United Kingdom and
Italy, which attempts to estimate the willingness to pay for conservation of 5% and 20% of
Amazonia41. This research is especially interesting as it is directed specifically towards the
Amazon and studies the WTP for two different sizes of protected areas.
When the area to be protected varied by 300%, the WTP rose by only between 34% and
39% (depending on the order of the questions). This behaviour is represented by the power
function with an exponent varying between 0.21 and 0.24. These numbers are located just in
the centre of the interval of possible values for the z constant in the species–area curve.
Thus, the empirical economic estimation is in harmony with the empirical ecological relation,
offering strong support to the hypothesis assumed, that the existence value of the forest is
related to the existence value of its component species.
As the willingness to pay estimated for two different area sizes by Horton et al. (2002)41 and
the species-area curve have the same behaviour in relation to the area of the forest, it is
feasible to assume that the WTP would maintain this behaviour for the entire area under
analysis.
Another advantage of the work by Horton et al. (2002)41 was that the authors asked about
the motivations for payment. As for 33% of respondents the main motive was concern for
future generations and for 17% concern was related to the preservation of species diversity
(but not for personal use, as on the option value), these percentages were used here to
19
calculate the total values in relation to bequest and existence values, as shown in the fifth
row of Table 4.
Following the methodology employed by Seroa da Motta (2005)17 these values for families in
the UK and Italy will be extrapolated to total global WTP, using PPP-adjusted per capita
income6 (sixth row).
And finally, from these total global values found, it is possible to estimate the equations that
show the behaviour of total bequest and existence values as a function of the area of the
forest (seventh row). From these equations, it is possible to calculate the equations that
govern the behaviour of the marginal values7 (eighth row).
Table 4.1:
Non-Use Values
Area (%)
Area (million ha)
Original WTP per family
WTP per family related to each
value
Global WTP related to each value
Bequest Value
5%
16.8
US$ 36
US$ 11,88
20%
67.2
US$ 50
US$ 16,5
Total Value Function
US$
4,3 US$
x109
x109
Y = 82.949.835x0,24
Marginal Value Function
y = 19.907.960x -0,76
Existence Value
5%
20%
16.8
67.2
US$ 36
US$ 50
US$ 6,12
US$ 8,5
5,9 US$
2,2 US$ 3,1 x109
9
x10
42.978.745x0,24
y = 10.314.899x -0,76
The equations that govern the marginal behaviour of the bequest and existence values are,
therefore, respectively:
Eq. 4.1: y = 19.907.960 x -0,76 | x E (1: 336.000.000)
Eq. 4.2: y = 10.314.899 x -0,76 | x E (1: 336.000.000)
Substituting the current forest area in the equations we find the values US$ 6.59 ha-1 yr-1 and
US$ 3.42 ha-1 yr-1, which represent, respectively, the marginal bequest value and the
marginal existence value at the present stage of the conversion process.
Due to the global character of the non-use values, supported by the fact that all the
estimations carried out here came from contingency valuation studies carried out on other
continents, its benefits will be added to the global portion of benefits of conservation of the
forest.
6
Using the formula WTPD=WTPO(PPPYD/PPPYO)e where e is the marginal elasticity of income (supposing
e=1), WTPO the willingness to pay in the original region, WTPD the willingness to pay for the region that one
wishes to adjust to, and PPPYD/PPPYO the ratio between income per capita (measured in purchase power
parity) of the destination region and the original one. Countries were grouped into three blocks, corresponding
to high, medium and low income.
7
Calculations were made from data obtained by Horton et al. (2002) when these ask first about the larger area.
As was observed by Seroa da Motta (2005), “for an estimate of existence value, the use of values with
decreasing scopes is more conservative, as the anchorage affects lower and not greater stocks”.
20
THE COSTS OF CONSERVATION
As mentioned previously, the cost of conserving the forest has two main components. The
most important one is the opportunity cost – the benefits that could be obtained by
converting the area to other uses. Among several studies reviewed that suggest values for
alternative land-uses, three recent ones17,18,42 focus specifically in the Brazilian Amazon and
follow careful albeit different methodologies.
Andersen et al. (2002)18 estimated the impact that the conversion of one additional hectare
had in the rural GNP of the Amazon region between 1985-1995. They concluded that one
additional converted hectare generated an increase of US$ 48,48 per year in the region’s
rural GDP.
Seroa da Motta (2005)17 estimated his opportunity costs based on the values for leasing land
for cattle ranching in the Amazon. He argues that the leasing annual value is much less
affected by the market failures that distort the land purchasing price. The values varied from
US$ 32,6 and US$ 49,5, with an average of US$ 38,4 ha-1 y-1.
Grieg-Gran (2006)101, on a paper commissioned by Stern (2006)42, estimates the return per
hectare for several land-uses in the Amazon region. Ponderating each one by the area it
currently occupies yields an annual value if US$ 57 ha-1 on the medium scenario.
The average between the three values can be considered a good estimate for the mix of
alternative uses available in the region.
It is then necessary to estimate the behavior of these benefits throughout the progress of
conversion. In opposition to the forest benefits’ case, there is no clear general dynamic
behavior here. It has been argued31 that these costs would drop when the remaining area
diminishes, as the areas with higher value for alternative uses would be converted first. It is
not clear, however, if this is the major driver behind the geographical distribution of the
conversion process. In the Brazilian Amazon, the transport infrastructure seems to play a
greater role46,47,102. As this infrastructure is expanded, the returns from conversion would
rise.
An important factor is the relative share that each alternative use has in the total mix of
future deforested areas. Although cattle ranching is still the dominant alternative land-use in
the Amazon region35, soybean plantation is becoming increasingly important37. Although the
majority (58%) of new soybean crops occupies already deforested land, a significant share
(42%) is direct conversion of forest103. FAO (2006)48 projects that world soybean production
will more than double until 2030 and more than triple until 2050, even when accounting for a
sharp decline in the market rate of growth. Brazil is forecast to overcome the US and
become the main exporter in 2007/08 and the main producer in 2012/13 [104]. As a result, it is
projected to expand the harvesting area from 21 to 31 million hectares until 2015/16 [104] and
to more then double its current area until 2050 [48]. These projections already suppose that
Brazil will increase its productivity from current 2,72 t/ha to 2,99t/ha, becoming more
productive than the US.
As returns from soybean plantation are nearly five times higher than returns from cattle
ranching101, the likely increase in the former’s share of new deforested areas would lead to
higher opportunity costs in the future. Despite the likely increase, another conservative
assumption will be adopted by considering that these costs will remain constant as the
conversion proceeds.
The second component is the cost involved in the maintenance and supervision of the
protected areas. The value chosen, US$ 2.36 ha-1 y-1, is estimated by James et al (1999)43
from smaller and fragmented areas, being probable that the cost per hectare of protecting a
21
much larger area would be considerably lower, due to the reduction in the perimeter/area
ratio and possible economies of scale. According to Pearce (2007)31, these costs would rise
as the conversion proceed. Data from James et al. (1999)43 suggest that they would fall. In
here they will be considered constant and so the marginal costs of conservation, totaling U$
50.36, will be constant throughout the conversion process.
Table 4.2:
Results
Spatial
Scale
Current
Marginal
Value
Marginal Function
Timber
National
US$ 0
 y = 33 | x ∈ (1;34.000.000)

 y = 0 | x ∈ (34.000.001;336.000.000)
Direct Use
NTFPs
Values
National
US$ 0
 y = 16.94 | x ∈ (1;8800000)

 y = 0 | x ∈ (8800001;336000000)
Recreation
National
US$ 0
 y = 76.63 | x ∈ (1;26000000)

 y = 0 | x ∈ (26000001;336000000)
Climate
Regulation
Global
US$ 43.13
y
=
14.491.680.000/x
(1:336,000,000)
Waste
Treatment
National
US$ 5.08
National
US$ 0.51
Erosion
Control
National
US$ 9.60
Water
Regulation
National
US$ 1.16
y = 389.760.000/x | x E (1;336.000.000)
National
US$ 0.56
y = 188.160.000/x | x E (1;336.000.000)
Global
US$ 10.26
y = 25.462.546
336.000.000)
Global
US$ 6.59
y = 19.907.960x -0,76 | x E (1:336.000.000)
Global
US$ 3.42
Disturbance
Indirect Use Regulation
Values
Option
Value
Bequest
Value
Existence
Value
Forest
Global
Benefits
Water
Provision
Option
Value
Bequest
Value
Existence
Value
US$ 80.31
22
|
x
E
y = 1.706.880.000/x | x E (1;336.000.000)
y= 171.360.000/x | x E (1;336.000.000)
y = 3.225.600.000/x | x E (1; 336.000.000)
x
-0,75
|
x
E
(1:
y = 10.314.899x -0,76 | x E (1:336.000.000)
Forest
US$ 18.53
National
Benefits
Benefits of
US$ 50.36
Conversion
Numbers in the fourth column are in ha-1 yr-1. Variable x represents the area of the Amazon
(in ha.). “Forest National Benefits” are the sum of all values of national scale plus Brazilian
share (2.55% – see text) of global values.
COST-BENEFIT ANALYSIS
Each service was classified according to its spatial nature into “national” or “global” values.
Brazil also absorbs part of the global benefits provided by the forest. Brazil’s share of gross
world product (PPP adjusted) is 2,55% and its share of world population is 2,89%. Albeit in
this case both figures are very similar, the first one is more economically sound (as it is
linked to potential WTP) and more fair. Therefore 2,55% of the global benefits will be
included in the national benefits.
In the first place, the present picture of the conversion process in the Brazilian Amazon
provides two observations. The first is that the current global benefits of conservation,
estimated at approximately US$ 80 ha-1 yr-1 are higher than the benefits of conversion,
estimated at approximately US$ 50 ha-1 yr-1. This indicates that the conversion process
practiced today in the Amazon is an inefficient activity, presenting costs greater than the
benefits. This conclusion was reached even with the conservative assumptions for the global
conservation values.
The second observation is that the sum of marginal benefits of national scale currently
provided by the forest, which is approximately US$ 18.5 ha-1 yr-1, is considerably lower than
the marginal benefits generated by alternative activities, the aforementioned US$ 50 ha-1 yr1
. Following the same rationale, the conclusion is that, from a national viewpoint, conversion
still brings more benefits than conservation.
These two observations are in perfect harmony with the observations obtained from the
analysis of comparative case studies made previously, demonstrating once again the
decisive role that global externalities have in this context.
As mentioned previously, this role becomes even clearer from a dynamic analysis of these
benefits. Figure 2 is an empirical version of Figure 1, showing the dynamic behaviour of the
national and global marginal benefits provided by the forest, as well as the marginal costs.
23
Figure 2
One can observe that the curve representing the global benefits of the forest, even with the
conservative assumptions applied, is constantly above the line of the costs of conservation.
Thus, the optimal point from the global point of view (Qglob) would be located to the right of
the current area, indicating that the conversion of the forest today already presents more
costs than benefits from the global standpoint.
On the other hand, it can be seen that the national benefits provided by the forest, initially
lower than those for conversion, become higher than these when conversion progresses
after the point Qnat. According to economic theory, this would be the point where the
process of conversion in the Brazilian Amazon would be interrupted should the Brazilian
government act in order to maximize national welfare. Solving the equation
(
)
5681760000 / x + 0.0255 1449168000 0/x + 25462546x -0.75 + 19.907.960 x -0.76 + 10.314.899 x -0.76 = 50.36
yields the national equilibrium point, approximately equal to 123 million hectares.
The dynamic analysis thus allows the identification of the national equilibrium point for the
process of conversion in The Amazon, which would occur at an area approximately equal to
36.6% of the current area (or 30.7% of the original extent).
Note that if the Brazilian government does not intervene, the process of conversion would
continue up to the optimum private point. Private owners are generally supposed to be
concerned only with direct use benefits. It is likely that sustainable timber extraction and
recreational activities conflict with each other to some extent. As a further conservative
assumption, the minimum area of both activities will be fully summed up. Notwithstanding,
the private optimum point would be located even further to the left of the national one,
corresponding to approximately 18% of the current area. Therefore it would be necessary to
correct national market failures so that the conversion process stabilizes at the efficient point
from the national point of view, or 36.6% of the current area, and does not proceed until the
private efficient point, equal to half of this area.
This correction, however, is not sufficient. At the present point the conversion process of
Brazilian Amazon has already become an undesirable activity that generates more losses
than benefits to humanity.
24
The dynamic approach adopted here allows visualization of the role that externalities play in
this context. This failure in the economic system, represented graphically by the distance
between the national and global marginal benefits curves, causes the optimal national and
global points not to coincide.
However, perhaps the most attractive aspect of this approach is yet to come. As the graph
above is a chart of marginal benefits, the area below each curve represents total benefits.
Suppose that the country holding these natural resources, in this case Brazil, wished to
interrupt the process of conversion immediately, considering that this would be the
preferable option for humanity as a whole. In this case, the area marked by the letter A
would represent the total economic incentives contrary to this alternative, equivalent to the
aggregate loss of welfare that Brazilian society would incur from this decision. These
incentives against conservation can be calculated from the equation:
336000000
=
∫
336000000
M arg inal Benefits of Conversion −
122855539
336000000
=
∫ 50.36 −
122855539
∫ M arg inal National Benefits of
Conservation =
122855539
336000000
∫ 5439840000 / x
(
)
+ 0.0255 14491680000/x + 25462546x -0.75 + 19.907.960x -0.76 + 10.314.899x -0.76 =
122855539
= US $ 4490893612
These annual US$ 4,5 billion would represent a considerable loss to national welfare for a
developing country as Brazil (and the vast majority of countries owning tropical forests), thus
being strong incentives against the option for conservation.
But the graph above offers some reasons for optimism. The gain in welfare that all the other
countries in the world would have from the immediate interruption of the conversion process
– the international externality –would be equal to the sums of areas A and B in the graph
above. That is, the benefit to the rest from the world of this decision is greater than the loss
incurred by the country. This shows a potential for win-win cooperation, with gains for both
sides.
If the rest of the world compensates the host country for a value equal to area A, its welfare
at the current point becomes equal to its welfare at point Qnat, allowing immediate
interruption of the conversion process without loss of national welfare.
The way in which this sum would be distributed internally is a question related to the specific
characteristics of each country. The important point here is that compensation for a value
equivalent to area A would counteract the economic incentives against conservation,
allowing interruption of the process without loss of national welfare. Addressing the problem
from a national level helps minimize the threat leakage, a recurring problem of conservation
programs. There would still be the threat of an international leakage, which could be
addressed through the extension of the cooperation to other countries hosting tropical
forests. Making annual payments subject to prior auditing of the state of the forests is also a
powerful tool to promote compliance.
EXTRAPOLATING AND BRINGING TO SCALE
Extrapolation of the compensation values found here for other countries that have tropical
forests is not strictly correct, as the evaluation should be made specifically for each region.
On the other hand, comparison with values found in the literature indicates that the values
25
estimated here are relatively close to the values for other countries and to those of
developing countries as a whole. Thus, an extrapolation, with reservations, of the values
found could be seen as a fair approximation.
Dividing the net annual loss of national welfare estimated by the area of the Brazilian
Amazon yields an average loss of US$ 13.4 ha-1 yr-1. Multiplying it by the 1.3 billion hectares
of tropical forests on the planet, the total necessary to compensate all the countries that
have tropical forests for immediate halting of the process of conversion would be equal to
US$ 17.4 billions per year. It is important to note that in contrast to several estimates in the
literature for the costs of conserving 10% or similar proportions, this figure relates to the
conservation of 100% of the tropical forest existent today.
These values cast down the criticism that a cooperation of this nature would be prohibitively
expensive. The estimated sum total for a global cooperation that would actually halt tropical
deforestation is equal to 0.04% of annual global income, to 4% of what rich countries
proportionally spend on their own internal environmental protection13, to 2% of what
governments currently spend on environmental harmful subsidies44 or to 2% of annual
military spending45. Actual willingness to pay for conservation of only 20% of the Amazon
forest has been estimated at US$ 20 billion a year41.
COMPARISONS WITH OTHER STUDIES
Although no study was found that made projections for the area of equilibrium of the
Brazilian Amazon, there are many projections for the area converted in a limited time
horizon. A recent study projected that 40% of the forest existing today will have been
converted by 2050[46]. Another projection suggests that the large scale infrastructure
expansion now underway in the Amazon will result in a forest loss of 28-42% as early as
2020[47]. The persistence and possible acceleration of the conversion process in the Amazon
is also consistent with projected land demand for agriculture expansion, as discussed
earlier48.
Comparison can also be made with the historical behavior of the conversion process across
the globe. According to the Millennium Ecosystem Assessment1, two ecosystem types
located in regions where the conversion process first begun (Mediterranean forests,
woodlands and scrubs and temperate forest steppe and woodland) have already lost c. 70%
of their extent. Of the other 8 productive ecosystem types subject to conversion, 5 will reach
similar levels by 2050 (temperate broadleaf and mixed forests; tropical and subtropical dry
broadleaf forests; flooded grasslands and savannas; tropical and subtropical grasslands,
savannas and shrublands; tropical and subtropical coniferous forests). Other study104
estimate that the extent of the frontier forests existent today corresponds to 21,7% of their
original extent.
The remaining stock of forests in all 72 countries (except Russia) for which the necessary
data was available the three regions where conversion has been active for a longer period is
also a good base for comparisons. Current regional forest cover corresponds to 33%
(Europe8), 8,4% (North Africa) and 21,5% (Continental Asia) of the original cover. In total,
the original forest cover amounted to 1.74 billion hectares. It has been reduced to 422 million
hectares, or slightly less than a quarter.
As for the monetary values necessary for compensation, the literature also offers opportunity
for comparisons. As discussed in the Research Highlights section, studies on this theme
usually suggests that the compensation value should cover the opportunity and
management costs of conservation. Few deduct from these the national benefits provided by
8
If Sweden, Norway and Finland are excluded, Europe’s current extent drops to 21% of the original cover.
26
the environment. And as shown in this study, this net conservation cost is still different
(probably higher) than the necessary compensation.
A precursor work by Cartwright49 based on the returns from timber exploitation suggests that
tropical countries would be unlikely to accept less then US$ 20 (or US$ 32 when indexed to
2002) as a compensation for not converting their forests. The net conservation costs
incurred by Kenya from its reserves were estimated21 at US$ 57 ha-1 y-1. In a global
simulation, the economic loss of setting aside reserves (opportunity costs) was estimated108
at US$ 40.5 ha-1 y-1 for the “Rest of the World” (which included Latin America). An
estimation based on empirical data for a global network of protected areas43 suggested that
the opportunity (based on land purchase price) and management costs of new areas in Latin
America would be US$ 9.4 ha-1 y-1. Kremen et al. (2000)19 estimated the net conservation
costs for a National Park in Madagascar at US$ 57 ha-1 y-1 (from the full capture scenario,
10% discount rate, 30 years timespan). The World Bank estimated109 the opportunity and
management costs of setting aside new areas as protected forest in developing countries at
US$ 93 ha-1y-1. A Payment for Ecosystem Services in Costa Rica, financed by the World
Bank, pays US$ 45 ha-1 y-1 to landowners who conserve forests110. Finally the three works
that based the estimation of the opportunity costs in this study offered values of US$ 38,4[17]
, US$ 48,5[18] and US$ 57[42].
Given the remarkably different methodologies and the spatial and temporal differences
between the ten estimates, the results are relatively close. Apart from the highest and the
lowest values, the other 8 studies are located in the US$ 32 – US$ 57 interval, with a mean
of US$ 46.9. Including the lowest and the highest values have a very small effect on the
mean (now at US$ 47.8). These mean values are close to the opportunity costs (US$50,4)
and are higher than the current net marginal conservation costs (US$ 32.5) and the average
compensation value (US$ 14.1) estimated here. The fact that the last two values are inside
the interval and below the mean indicates that the quantitative values estimated are not
overly distorted and may be considered conservative estimates.
CONSISTENCY AND LIMITATIONS
Due to the many theoretical and practical obstacles found, it was necessary to adopt some
simplifications. Some of the consequent limitations are:
1)The Brazilian Amazon is not a single, homogeneous forest block. There are considerable
regional differences in several ecological characteristics (i.e. soil composition, species
diversity, climatic regimes) that impact the benefits provided by the forest. In addition to
being impacted by these ecological differences, the benefits provided by alternative uses are
also highly influenced by regional differences in socio-economic characteristics (i.e. transport
infrastructure, distance to markets).
2) In order to analyse the benefits provided by the ecosystems to mankind, those have been
classified into different “services” and analysed individually. Although this is the common
practice in the literature, its is important to keep in mind that ecological systems are complex
net of interactions111 and that alterations in one service impacts the provision of others. For
instance, species extinction is likely to impact the climate regulation service38, while
interactions between the latter and the hydrological cycle have recently been better
understood39,40. These feedbacks often present time lags, adding more complexity to the
analysis. Further research that introduces these interrelations into the analysis can help to
reduce distortion in the results.
3) Although a large portion of the key values were obtained from studies focused in the
Brazilian Amazon, estimates from other tropical countries were also used. Despite being a
common practice in the literature, the issue of transferring environmental benefits estimates
across different countries is a very complex one. Different environmental, economic and
27
socio-cultural characteristics imply different values for each country. Although some tools
were used to reduce the distortions, this is still a source of error.
4) Apart from timber, where all the data necessary for the estimations was available, the
simplifying assumptions adopted in the estimations of the values related to the other direct
use benefits of the forest probably distorted the individual results. These distortions,
however, would only begin to influence the cost-benefit analysis if the largest of the minimum
areas found were more than doubled. Given the conservative assumptions adopted, it is
unlikely that these minimum areas have been so strongly underestimated.
5) The estimation of some of the indirect use values is probably the major limitation in this
study. Apart from the climate regulation service, whose marginal benefits were estimated
directly, the indirect marginal benefits were obtained through the conversion of published
average values using the rectangular hyperbolic function. Although based on a published
work and presenting the dynamic behavior expected by several different authors, the choice
of this functions it is still an arbitrary one. There are other functions that would also match
the expected behavior and produce different quantitative results. It would even be possible
to select a function that would result in the national benefits from the forest being higher than
the benefits of alternative uses. This choice, however, in addition to being as arbitrary as the
one made here, would lead to conclusions that contradict the ones obtained from the review
of case studies carried earlier. It is this consensus found in the analysis of the literature that
gives a qualitative support to the conclusions found here.
6) The marginal option value estimated should be considered a conservative lower-limit
bound, as the potential for pharmacological application is only a portion of the new benefits
that can arise from the conserved forest.
7) The value chosen to reflect the benefits from the conversion to alternative uses was
based in two recent and methodologically strong analysis focused on the Brazilian Amazon.
Its dynamic behavior, however, was not the focus of this study and the supposition adopted
here that it would remain constant throughout the conversion process is a very simplistic
one. According to projected market estimates, however, it is likely that this is a conservative
estimate and a more realistic approach would reinforce the qualitative conclusions found.
8) Finally, in addition to these, this work also carries all the limitations expressed by the
authors of the studies whose estimates were used.
Overall, all the simplifications certainly have a quantitative impact on the results that,
therefore, should be used with caution. The impact could certainly be reduced with additional
research. Although this quantitative distortion certainly exists, the subsequent critical
discussion and comparison with similar values in the literature indicate that it may not be
overly large. And, more importantly, the analysis of comparative case studies made
previously shows that it is unlikely that these distortions have had a qualitative influence on
the main conclusions of this study.
RESEARCH INSIGHTS
Even with reservations about some quantitative estimates, the dynamic marginal approach
adopted provided important research insights, that can explored in future:
1)Recognizing the different characteristics of each category of the TEV and using different
approaches for each one allowed to overcome some of the pitfalls encountered in the
analysis.
2)There are some misinterpretations in the literature about the nature of the values
estimated. Estimating values for one unit (usually a hectare) of a natural area does not mean
it is marginal value. A unit value can be an average (usually obtained by dividing a value
found for the whole ecosystem by its area), a maximum (usually obtained by estimating the
potential value that could be produced in a unit) or a marginal value, which is how much one
28
more unit contributes to the total value. The last one is usually more useful for policy
purposes.
3) The focus on a very extensive area brought up an interesting scale issue in relation to the
marginal direct use values. The vast majority of works in the literature estimate values for
much smaller areas, such as specific reserves and parks. It is generally assumed that there
will be a market for all the goods potentially produced in the whole area, yielding a positive
marginal value. That may be true for small areas. When we start “zooming out” the picture,
however, this assumption ceases to be valid after a certain point and the marginal value
drops to zero.
4) Due to the valuation process used, estimations for indirect, option and non-use values
generally yield values of average nature. If the general behaviour of marginal values is the
one depicted in Figure 1 – as the literature suggests – then these average values can be
considerably different than current marginal values.
5)Applying an empirical ecological function into the economic valuation process allowed the
estimation of a sound dynamic behaviour for option, bequest and existence values. In the
case of the latter two, the approach was supported and confirmed by empirical economic
estimates. This highlights the still largely untapped potential of cooperation between these
two sciences.
6)Works in the literature addressing a cooperation such as the one proposed here usually
argue that the payment by the international community should be equal to the opportunity
and management costs incurred by the host nation. In fact the national benefits of the
conserved area should be deducted from those costs (to find the incremental costs). The
dynamic analysis also shows that simply deducting current marginal national benefits from
current marginal costs and multiplying them by the area in order to find total compensation
may not be strictly correct. Supposing that the national benefits would rise faster than the
costs (the only way to have Qnat>0), this procedure would overestimate the necessary
compensation. As an illustration, current marginal costs of conservation were estimated here
at US$ 50.4 ha-1 yr-1, current net marginal costs at US$ 31.8 ha-1 yr-1 and the average
compensation costs at US$ 13.4 ha-1 yr-1.
THE ROLE OF THE INTERNATIONAL EXTERNALITY
The initial analysis of comparative case studies hinted on the decisive role that international
externalities have in the persistence of the conversion process in lower-income countries.
This classic market failure arises in this case from the interaction of the different spatial
dimensions of ecosystem services with the division of the planet in sovereign sub-units. The
consequence is that a country does not receive all the benefits arising from its ecosystems
and, when bearing all the costs of their conservation, is attracted to a national point of
equilibrium which is lower than the global optimum. The role of the externality becomes clear
in the graphic theoretical analysis (Fig. 1), which is also in perfect harmony with past and
current trends of the conversion process in both high and low income countries.
The impact of this market failure became frightening clear when the marginal dynamic
approach estimated that 63.4% of the current area of the Brazilian Amazon will be
deforested until it reaches the national equilibrium point. This proportion is consistent with
the past behaviour of the conversion process throughout the planet. Tropical deforestation is
unlikely to stop if the substantial economic incentives contrary to conservation – here
estimated at annual US$ 4.5 billion for the Brazilian Amazon and US$ 17.4 billion for all
tropical forests – are not counterbalanced.
29
It is important to understand that these incentives do influence private and governmental
decisions even if they are not explicitly accounted for or even directly perceived. Their
diffuse force acts through the political process, resulting in governmental decisions to
“develop” natural environments. Even when governments resist these pressures, maybe
against the interests21,31 of their own people, the invisible hand pushes national and foreign
private groups to illegal acts, whose repression demands additional sacrifice of public
resources.
In addition to being unrealistic, it is hardly fair to demand that the population of tropical
countries sacrifice part of their scarce well-being by forfeiting their “national optimum path”,
in order to provide the global optimum amount of ecosystem services to an international
community, of which many members already enjoy high levels of affluence. It is fairer to
expect that this community transfer back a small part of the global benefits these
ecosystems provide in order to compensate their host countries by the net loss incurred in
conserving them at global optimum levels.
DISCUSSION
The idea of an international compensation mechanism against tropical deforestation is not
new17,18,26,42,43,49 and is gaining momentum. Following proposals by tropical countries in the
2005 and 2006 annual Conferences (cit), the UN Convention on Climate Change is currently
studying the issue, with a report due in the 2007 Conference50. A similar suggestion was
also made by the most complete analysis on the economics of climate change to date,
where avoiding tropical deforestation was pointed out as one of the cheapest and most
straightforward ways of combating climate change42.
The fact that both references are related to the climate issue should not, however, lead to
the inference that a cooperation as proposed here and elsewhere should necessarily be
achieved exclusively through existing climate change frameworks. Including avoided
deforestation in the general carbon market has still unsolved technical issues and potential
destabilizing consequences. In addition, since here the main stakeholders form a smaller
group, with ten countries hosting 75% of the world’s tropical forests4 and the EU, the US,
Japan and Canada holding 70% of the global wealth, achieving cooperation should be
easier.
Certainly complex political negotiations on national and international levels are part of the
way forward. Nonetheless, there is room for optimism. The analysis performed here, even
whilst ignoring a wide range of unquantified positive environmental and humanitarian “sideeffects” and being based on the selfish utility of individuals, showed that a cooperation is a
strong win-win solution, with its global benefits outweighing by far the amount of
compensation necessary.
30
References
1.
Millennium Ecosystem Assessment Ecosystems and Human Well-being: Synthesis
(Island Press, Washington DC, 2005).
2.
Vitousek, P. M., Mooney, H. A., Lubchenco, J. & Melillo, J. M. Human Domination of
Earth's Ecosystems. Science 277, 494-499 (1997).
3.
Haberl, H. The global socioeconomic energetic metabolism as a sustainability problem.
Energy 31, 87-99 (2006).
4.
Food and Agriculture Organization Global Forest Resource Assessment 2000 (FAO,
Rome, 2001).
5.
Field, C. B., Behrenfeld, M. J., Randerson, J. T., Falkowski, P. Primary Production of
the Biosphere: Integrating Terrestrial and Oceanic Components. Science 281,
237-240 (1998).
6.
Wilson. E. O.(ed) Biodiversity (National Academy Press, Washington DC, 1988).
7.
Achard, F. et al. Determination of Deforestation Rates of the World’s Humid Tropical
Forests. Science 297, 999-1002 (2002).
8.
Baumert, K. A., Herzog, T. & Pershing, J. Navigating the Numbers: Greenhouse Gas
Data and International Climate Policy (World Resources Institute, Washington
DC, 2005).
9.
Houghton, R . A . Aboveground Forest Biomass and the Global Carbon Balance. Glob.
Change Bio. 11, 945–958 (2005).
10. Sala, O. E. et al. Global Biodiversity Scenarios for the Year 2100. Science 287, 17701774 (2000).
11. Brook, B. W., Sodhi, N. S. & Ng, P. K. L. Catastrophic extinctions follow deforestation
in Singapore. Nature 424, 420-423 (2003).
12. Daily, G. (ed.) Nature’s Services: Societal Dependence on Natural Ecosystems (Island,
Washigton DC, 1997).
13. Pearce, D. New directions for financing global environmental change. Global
Environmental Change 5, 27- 40 (1995).
14. Pinedo-Vasques, M., Zarin, D. & Jipp, P. Economic returns from forest conversion in
the Peruvian Amazon. Ecol. Econ. 6, 76–78 (1992).
15. Yaron, G. Forest Plantation crops or small-scale agriculture? An economic analysis of
alternative land use options in the Mount Cameroon Area. Journal of
Environmental Planning and Management 44, 85-108 (2001).
16. Kumari, K., Sustainable Forest Management in Peninsular Malaysia: Towards a total
economic valuation approach (Thesis, University of East Anglia, Norwich, UK,
1994).
17. Seroa Da Motta, R. Custos e Beneficios do desmatamento na Amazônia. Ciencia e
Ambiente 32, 73-84 (2005).
18. Andersen, L. A., Granger, O., Reis, E. J., Weinhold, D., Wunder, S. The Dynamics of
Deforestation and Economic Growth in the Brazilian Amazon (Cambridge
University Press, Cambridge, 2002).
19. Kremen, C. et al. Economic incentives for rain forest conservation across scales.
Science 288, 1828-1832 (2000).
20. Batagoda, S. Turner, R.K. Tinch, R. & Brown, K. Towards policy relevant ecosystem
services and natural capital values: rainforest non-timber products. CSERGE
Working Paper GEC 2000-06 (2000).
21. Norton-Griffiths, M. & Southey, C. The opportunity costs of biodiversity conservation in
Kenya. Ecol. Econ. 12, 125-139 (1995).
22. Sathirathai, S., Economic valuation of mangroves and the roles of local communities in
the conservation of the resources: case study of Surat Thani, South of Thailand.
Final report submitted to the Economy and Environment Program for Southest
Asia (EEPSEA), Singapore, 1998.
23. Garrod, G.D. & Willis, K.G. Valuing goods’ characteristics: An application of the
hedonic price method to environmental attributes. Journal of Environmental
Management 34, 59–61 (1992).
31
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
Carson, R.T., Wilks, L. & Imber, D. Valuing the Preservation of Australia's Kakadu
C4nservation Zone. Oxford Economic Papers 46, 727-749 (1994).
Van Vuuren W. & Roy, P. Private and Social Returns from Wetland Preservation
Versus Those From Wetland Conversion to Agriculture. Ecol. Econ. 8, 289-305
(1993).
Turner, R.K. et al. Valuing nature: lessons learned and future research directions. Ecol.
Econ. 46, 493-510 (2003).
Balmford, A. et al. Economic Reasons for Conserving Wild Nature. Science 297, 950953 (2002).
Costanza, R. et al. The value of the world’s ecosystem services and natural capital.
Nature 387, 253-260 (1997).
Torras, M. The total economic value of Amazonia deforestation, 1978-1993. Ecol.
Econ. 33, 283-297 (2000).
Farber, S.C., Costanza, R. & Wilson, M.A. Economic and ecological concepts for
valuing ecosystem services. Ecol. Econ. 41, 375-392 (2002).
Pearce, D.W. Do we really care about biodiversity ?. Environmental and Resource
Economics, forthcoming (2007).
Holling, C.S. Resilience and stability of ecological systems. Annual Review of Ecology
and Systematics 4, 1-23 (1973).
Perrings,C. & Pearce, D.W. Threshold Effects and Incentives for the Conservation of
Biodiversity. Environmental and Resource Economics 4, 13-28 (1994).
Perrings,C. Biodiversity conservation as insurance, in Swanson, T. M. (Ed.) The
Economics and Ecology of Biodiversity Decline (Cambridge University Press,
Cambridge,1995).
Fearnside, P. M. Deforestation in Brazilian Amazonia: History, Rates and
Consequences. Conservation Biology 19, 680–688 (2005).
Asner, G.P. et al. Selective Logging in the Brazilian Amazon. Science 310, 480-482
(2005).
Morton, D.C. et al, Cropland expansion changes deforestation dynamics in the
southern Brazilian Amazon. Proc. Natl. Acad. Sci. U.S.A. 103, 14637–14641
(2006).
Bunker, D. E. et al. Species Loss and Aboveground Carbon Storage in a Tropical
Forest. Science 310, 1029-1031 (2005).
Richey, J. E., Melack, J. M., Aufdenkampe, A. K., Ballester, M. V. & Hess, L. L.
Outgassing from Amazonian rivers and wetlands as a large tropical source
ofatmospheric CO2. Nature 416, 617–620 (2002).
Mayorga, E. et al. Young organic matter as a source of carbon dioxide outgassing from
Amazonian rivers. Nature 436, 538-541 (2005).
Horton, B. Collarullo, G. Bateman, I.J. & Peres, C. Evaluating non-user willingness to
pay for a large scale conservation program in Amazonia: A UK/Italian contingent
valuation study. Environmental Conservation 30, 139-146 (2003).
Stern, N. Stern Review on the Economics of Climate Change (Cambridge University
Press, Cambridge, 2006).
James, A. N., Gaston, K. J. & Balmford, A. Balancing the Earth’s accounts. Nature
401, 323-324 (1999).
Myers, N. Lifting the veil on perverse subsidies. Nature 392, 327–328 (1998).
SIPRI, Military Expenditure Database. Available electronically at www.sipri.org (2005).
Soares-Filho, B. S. et al. Modelling conservation in the Amazon basin. Nature 440,
520-523 (2006).
Laurance, W. F. et al. The Future of the Brazilian Amazon. Science 291, 438-439
(2001).
Food and Agriculture Organization World agriculture: towards 2030/2050 (FAO, Rome,
2006).
Cartwright, J. The politics of preserving natural areas in Third World States. The
Environmentalist 5, 179-186 (1985).
32
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
United Nations Framework Convention on Climate Change Reducing Emissions from
Deforestation
in
Developing
Countries.
Available
eletronically
at
http://unfccc.int/methods_and_science/lulucf/items/1084.php (2006).
De Groot, R. S., Wilson, M. A., Boumans, R. M. J. A typology for the classification,
description and valuation of ecosystem functions, goods and services. Ecological
Economics 41, 393-408 (2002).
Almeida, O. T. & Uhl, C. Identificando os custos de usos alternativos do solo para o
planejamento municipal da Amazônia: o caso de Paragominas (PA), in May, P.
(ed.) Economia Ecológica (Campus, Rio de Janeiro, 1995).
Nepstad, D. C. et al. Large-Scale Impoverishment of Amazonian Forests by Logging
and Fire. Nature 398, 505-508 (1999).
Lentini, M., Pereira, D., Celentano, D., Pereira, R. Fatos Florestais da Amazônia 2005
(Imazon, Belém, 2005).
IBGE, Produção da Extração Vegetal e da Sivicultura 2005 (IBGE, Rio de Janeiro,
2006).
World Bank, Global commodity Market (The World Bank, Washington D.C., 2000).
Peters, C.M., Gentry A.H. & Mendelsohn, R.O. Valuation of an Amazonian Rain
Forest. Nature 339, 655-656 (1989).
Anderson, A.B., May, P.H. & Balick, M. The Subsidy from Nature: Palm Forests,
Peasantry, and Development on an Amazon Frontier (Columbia University Press,
New York, 1991).
Chopra, K. The value of non-timber forest products: an estimation for tropical
deciduous forests in India. Economic Botany 47, 251-57 (1993).
Godoy, R., Lubowski, R. & Markandya, A. A method for the economic valuation of nontimber tropical forest products. Economic Botany 47, 220-233 (1993).
Grimes, A. et al. Valuing the Rain Forest: the economic value of nontimber forest
products in Ecuador. Ambio 23, 405-410 (1994).
Shone, B. M. & Caviglia-Harris, J. L. Quantifying and comparing the value of nontimber forest products in the Amazon. Ecol. Econ. 58, 249– 267(2006).
Tobias, D. & Mendelsohn, R. Valuing ecotourism in a tropical rain-forest reserve.
Ambio 20, 91-93 (1991).
Edwards, S. F. The demand for Galapagos vacations: estimation and application to
wilderness preservation. Coastal Management 19, 155-199 (1991).
Ruitenbeek, H.J. The rainforest supply price: a tool for evaluating rainforest
conservation expenditures. Ecol. Econ. 6, 57–78 (1992).
Houghton R.A., Lawrence K.T., Hackler J.L. & Brown, S. The spatial distribution of
forest biomass in the Brazilian Amazon: a comparison of estimates. Global
Change Biology 7, 731–746 (2001).
Fearnside, P.M. Greenhouse gases from deforestation in Brazilian Amazonia: net
committed emissions. Climatic Change 35, 321-360 (1997).
Nepstad, D.C. et al. ‘The Role of Deep Roots in the Hydrological and Carbon Cycles
of Amazonian Forests and Pastures. Nature 372, 666–669 (1994).
Asner, G. P., Keller, M. & Silva J. N. M. Spatial and temporal dynamics of forest
canopy gaps following selective logging in the eastern Amazon. Global Change
Biology 10, 765-783 (2004).
Achard F, Eva HD, Mayaux P., Stibig, H. & Belward, A. Improved estimates of net
carbon emissions from land cover change in the tropics for the 1990s. Global
Biogeochemical Cycles 18, GB2008 (2004).
Houghton, R.A. et al. Annual Fluxes of carbon from deforestation and regrowth in the
Brazilian Amazon. Nature 403, 301-304 (2000).
Fearnside, P. M. & Laurance, W. F. Tropical deforestation and greenhouse-gas
emissions. Ecological Applications 14, 982–986 (2004).
Fearnside, P. M. Global warming and tropical land-use change: greenhouse gas
emissions from biomass burning, decomposition and soils in forest conversion,
33
74.
75.
76.
77.
78.
79.
80.
81.
82.
83.
84.
85
86.
87.
88.
89.
90.
91.
92.
93.
94.
95.
shifting cultivation and secondary vegetation. Climatic Change 46, 115–158
(2000).
Schiermeier, Q. Market Watch. Nature 444, 533 - 533 (2006).
Clarkson, R. & Deyes, K. Estimating the Social Cost of Carbon Emissions. Government
Economic Service Working Paper (HM Treasury, London, 2002).
Raich, J. W. Russell, A. E., Kitayama, K., Parton, W. J Vitousek, P. M. Temperature
influences carbon accumulation in moist tropical forests. Ecology 87, 76-87
(2006).
Brown, S. and Lugo, A. E. Aboveground Biomass Estimates for Tropical Moist Forests
of the Brazilian Amazon. Interciencia 17, 8–18 (1992).
Potter, C. S., Terrestrial Biomass and the Effects of Deforestation on the Global
Carbon Cycle BioScience 49, 769-778 (1999).
Pearce, D. An economic approach to saving the tropical forests in Helm, D. (ed.)
Economic policy toward the environment (Blackwell, Oxford, 1991).
Brown, K. & Pearce, D. W. The economic value of non-marketed benefits of tropical
forests: carbon storage in: Weiss, J. (ed) The economics of project appraisal and
the environment (Edward Elgar, London, 1994).
Pearce, D. et al. The social costs of climate change: Greenhouse damage and the
benefits of control in IPCC Climate Change 1995: Economic and Social
Dimensions of Climate Change (Cambridge University Press, Cambridge, 1996).
Fearnside, P. Environmental services as a strategy for sustainable development in
rural Amazonia. Ecol. Econ. 20, 53–70 (1997).
Pearce, D.W. The social cost of carbon and its policy implications. Oxford Rev. Econ.
Policy 19, 362–384 (2003).
Pearce, D., and Moran, D. The economic value of biodiversity (Earthscan Publication,
London, 1994).
Krutilla, J. V.. Environmental resource services of Malaysian moist tropical forest
(Johns Hopkins University Press, for Resources for the Future, Inc, Baltimore,
1991).
Pimentel, D. et al. Economic and environmental benefits of biodiversity. Bio-Science
47, 747–757 (1997).
Cruz, W., Francisco, H.A. & Conway, Z.T The on-site and downstream costs of soil
erosion in the Magat and Pantabangan watersheds. Journal of Philippine
Development 26, 85-11 (1988).
Magrath, W. & Arens, P. The costs of soil erosion on Java: a natural resource
accounting approach. Environment Department Working Paper No. 18 (The
World Bank, Washington DC, 1989).
Dixon, J. A. & Hodgson, G. Economic evaluation of Coastal Resources: the El Niño
study. Tropical Coastal Area Management (August), 5 -7 (1988).
Chomitz, K.M. & Kumari, K. The Domestic Benefits of Tropical Forests: A Critical
Review Emphasizing Hydrological Functions. World Bank Policy Research
Working Paper No. 1601 (The World Bank , PRDEI, Washington, DC, 1996).
Food and Agriculture Organization Irrigation water use per country in the year 2000.
Available electronically at http://www.fao.org/landandwater/aglw/aquastat/dbase/
index.stm (2006).
Food and Agriculture Organization State of the World’s Forests 2003 (FAO, Rome,
2003).
Whitehead, P.G. & Robinson, M. Experimental basin studies - an international and
historical perspective of forest impacts. Journal of Hydrology 145, 217-230
(1993).
Simpson, R. D., Sedjo, R. A. & Reid, J. W. Valuing Biodiversity for Use in
Pharmaceutical Research. Journal of Political Economy 104, 163-85 (1996).
Rausser, G. C. & Small, A. A. Valuing Research Leads: Biodiversity Prospecting and
the Conservation of Genetic Resources. Journal of Political Economy 108, 173206 (2000).
34
96.
97.
98.
99.
100.
101.
102.
103.
104.
105.
106.
107.
108.
109.
110.
111.
Connor, E. F., & Mccoy, E. D. Statistics and biology of the species-area relationship.
Am. Nat. 113,791-833 (1979).
Reid, W. R. How many species will there be in Whitmore, T. C. & Sayer, J. A. (ed.)
Tropical deforestation and species extinction (Chapman and Hall, New York,
1992).
Pearce, D. W. & Whittington G. Economic values and the environment in developing
world. United Nations Enviroment Programme by CSERGE and University of
North Carolina at Chapel Hill (1994).
Adger, W.N., Brown, K., Cervigni, R. & Moran, D. Total economic value of forests in
Mexico. Ambio 24, 286-296 (1995).
Mendelsohn, R. Notes on economic plants. Econ. Botany 51, 328 (1997).
Grieg-Gran, M. The Cost of Avoiding Deforestation. Report prepared for the Stern
Review of the Economics of Climate Change. Available electronically at
http://www.hmtreasury.gov.uk/media/8A7/17/stern_review_supporting_technical_m_greiggran_
261006a.pdf (2006).
Kirby, K.R. et al. The future of deforestation in the Brazilian Amazon. Futures 38, 432453 (2006).
Brown, J. C. Koeppe, M. Coles, B. & Price, K. P. Soybean Production and Conversion
of Tropical Forest in the Brazilian Amazon: The Case of Vilhena, Rondonia.
Ambio 34, 462-269 (2005).
FAPRI U.S. and World Agricultural Outlook 2006. Food and Agricultural Policy
Research Institute, Iowa State University. Available electronically at
http://www.fapri.iastate.edu/outlook2006/ (2006).
Bryant, D., Nielsen, D. & Tangley, L. The Last Frontier Forests: Ecosystems and
Economies on the Edge (World Resources Institute, Washington DC, 1997).
Food and Agriculture Organization FAOSTAT Online Statistical Service. Rome: FAO.
Available online at: http://faostat.fao.org (2006).
Billington,C. Kapos, V. Edwards, M. Blyth, S. & Iremonger, S. Estimated Original
Forest Cover Map - A First Attempt (WCMC, Cambridge, 1996).
Lewandrowski, J, Darwin, R. F., Tsisgas, M. Raneses, A. Estimating costs of protecting
global ecosystem diversity. Ecological Economics 29, 111-125 (1999).
World Bank Costing the 7th Millennium Goal: Ensure Environmental Sustainability.
Environment Department and Development Economics Research Group, The
World Bank, Washigton DC (2000) [Restricted].
Pagiola, S. et al. Paying for biodiversity conservation services in agricultural
landscapes (Environment Department Paper n° 96, The World Bank, Washigton
DC, 2004).
Levin, S. A. Ecosystems and the Biosphere as Complex Adaptive Systems.
Ecosystems 1, 431-436 (1998).
35